Skip to content

Instantly share code, notes, and snippets.

View leigh-johnson's full-sized avatar
💜

Leigh Johnson Zaman leigh-johnson

💜
164 followers · 142 following
View GitHub Profile
@leigh-johnson
leigh-johnson / postprocessed-tflite.sh
Last active November 27, 2019 05:34
http://download.tensorflow.org/models/object_detection/ssd_mobilenet_v3_small_coco_2019_08_14.tar.gz
#!/usr/bin/env bash
TMP_DIR=${HOME}/tmp
TF_MODELS_DIR=${HOME}/projects/models
TF_MODELS_PY=${HOME}/projects/models/.venv/bin/python
MODEL_NAME=ssd_mobilenet_v3_small_coco_2019_08_14
MODEL_URL=http://download.tensorflow.org/models/object_detection/${MODEL_NAME}.tar.gz
MODEL_DIR=${TMP_DIR}/${MODEL_NAME}
CONFIG_FILE=${MODEL_DIR}/pipeline.config
CHECKPOINT=${MODEL_DIR}/model.ckpt
@leigh-johnson
leigh-johnson / mobilenet_v2_architechture.txt
Created July 7, 2019 04:32
Portable Computer Vision, SF Python July 2019
Model: "mobilenetv2_1.00_224"
__________________________________________________________________________________________________
Layer (type) Output Shape Param # Connected to
==================================================================================================
input_1 (InputLayer) [(None, 224, 224, 3) 0
__________________________________________________________________________________________________
Conv1_pad (ZeroPadding2D) (None, 225, 225, 3) 0 input_1[0][0]
__________________________________________________________________________________________________
Conv1 (Conv2D) (None, 112, 112, 32) 864 Conv1_pad[0][0]
__________________________________________________________________________________________________
@leigh-johnson
leigh-johnson / keras_lstm.py
Created July 6, 2019 23:28
Comparsing TensorFlow vs Keras for an LSTM model
from keras.layers import LSTM
model = Sequential()
model.add(Embedding(max_features, 256, input_length=maxlen))
model.add(LSTM(output_dim=128, activation='sigmoid', inner_activation='hard_sigmoid'))
model.add(Dropout(0.5)) model.add(Dense(1))
model.add(Activation('sigmoid'))
model.compile(loss='binary_crossentropy', optimizer='rmsprop', metrics=['accuracy'])
model.fit(X_train, Y_train, batch_size=16, nb_epoch=10)
@leigh-johnson
leigh-johnson / mobilenetv2_base_summary.txt
Created June 24, 2019 18:47
Model: "mobilenetv2_1.00_224"
__________________________________________________________________________________________________
Layer (type) Output Shape Param # Connected to
==================================================================================================
input_1 (InputLayer) [(None, 224, 224, 3) 0
__________________________________________________________________________________________________
Conv1_pad (ZeroPadding2D) (None, 225, 225, 3) 0 input_1[0][0]
__________________________________________________________________________________________________
Conv1 (Conv2D) (None, 112, 112, 32) 864 Conv1_pad[0][0]
__________________________________________________________________________________________________
@leigh-johnson
leigh-johnson / v2.0.0-beta0-build.log
Created June 15, 2019 09:32
Internal Compiler Error cross-compiling v2.0.0-beta0 for Raspberry Pi
This file has been truncated, but you can view the full file.
WORKSPACE: /Users/ljohnson/repos/tensorflow
CI_DOCKER_BUILD_EXTRA_PARAMS:
CI_DOCKER_EXTRA_PARAMS: -e CI_BUILD_PYTHON=python3 -e CROSSTOOL_PYTHON_INCLUDE_PATH=/usr/include/python3.4
COMMAND: tensorflow/tools/ci_build/pi/build_raspberry_pi.sh
CI_COMMAND_PREFIX: ./tensorflow/tools/ci_build/builds/with_the_same_user ./tensorflow/tools/ci_build/builds/configured pi-python3
CONTAINER_TYPE: pi-python3
BUILD_TAG: tf_ci
(docker container name will be tf_ci.pi-python3)
Building container (tf_ci.pi-python3)...
@leigh-johnson
leigh-johnson / armv7.build_history000
Last active June 12, 2019 05:28
Cross-compiling Tensorflow 2.0 for RPI
CI_DOCKER_EXTRA_PARAMS="-e CI_BUILD_PYTHON=python3 -e CROSSTOOL_PYTHON_INCLUDE_PATH=/usr/include/python3.5" \
tensorflow/tools/ci_build/ci_build.sh PI-PYTHON3 \
tensorflow/tools/ci_build/pi/build_raspberry_pi.sh
CC_TOOL=/tools/cross-pi-gcc-8.3.0-1/bin/arm-linux-gnueabihf-g++ \
bazel build -c opt \
--copt=-march=armv7-a \
--copt=-mfpu=neon-vfpv4 \
--copt=-std=c++11 \
--copt=-funsafe-math-optimizations --copt=-ftree-vectorize \
@leigh-johnson
leigh-johnson / install-go.sh
Last active November 19, 2018 07:56
Updated init scripts for kubeflow + microk8s
echo "Downloading and installing go"
wget https://dl.google.com/go/go1.11.2.linux-amd64.tar.gz /home/multipass/go1.11.2.linux-amd64.tar.gz
tar -xvf go1.11.2.linux-amd64.tar.gz
@leigh-johnson
leigh-johnson / gcloud_automl_demo.ipynb
Created September 9, 2018 05:10
Loading
Sorry, something went wrong. Reload?
Sorry, we cannot display this file.
Sorry, this file is invalid so it cannot be displayed.
@leigh-johnson
leigh-johnson / gcloud_automl_demo.ipynb
Created September 7, 2018 21:11
Loading
Sorry, something went wrong. Reload?
Sorry, we cannot display this file.
Sorry, this file is invalid so it cannot be displayed.
@leigh-johnson
leigh-johnson / nth_iterinterleave.py
Last active August 31, 2018 16:40
def nth_iterinterleave(*arrays, nth=1):
"""Interleave multiple lists.
via https://github.com/dgilland/pydash/blob/develop/src/pydash/arrays.py
Expanded to allow an nth parameter, where arrays[0] is treated as the primary and arrays[1:] are interleaved every n elements
Example:
given *arrays = [ [1,2,3,4,5,6,7,8,9,10], ['x', 'y', 'z'], ['a', 'b', 'c'] ]